HP 2013-2014: Explo Dive data
Philippine Chambault
2022-08-23
Background: High-resolution harbour porpoises data (dive data sampling rate: 1 sec from MK10 tags, plus low resolution Argos locations) collected from five individuals instrumented from 2013 to 2014 in Maniitsoq, West Greenland. Prey items found in HP West Greenland (Maniitsoq, Nuuk and Paamiut) according to Heide-Jorgensen et al 2011: Atlantic cod was detected in 31% and Greenland cod in 19% of the stomachs. Also: Mallotus villosus (Capelin, 78%), Boregadus saida (Atlantic cod, 53%), Gadidae (cods family, 59%), squid beaks or eyes (70%).
1 Import data
1.1 Dive data
##
## 22849b 22850b 27262 27262b 93100
## 12428775 11490281 1267945 11389946 139144131.2 Locations data
##
## 22849b 22850b 27262 27262b 93100
## 1210 1259 182 1136 581## # A tibble: 5 × 5
## id start end duration nlocs
## <chr> <dttm> <dttm> <drtn> <int>
## 1 22849b 2014-07-22 15:16:19 2014-12-15 13:43:53 145.93581 days 1210
## 2 22850b 2014-07-17 19:12:20 2014-12-01 13:33:26 136.76465 days 1259
## 3 27262 2013-10-02 15:53:13 2013-10-18 13:42:40 15.90934 days 182
## 4 27262b 2014-07-22 15:26:13 2014-12-09 13:23:16 139.91462 days 1136
## 5 93100 2014-07-22 14:05:17 2014-12-25 21:56:44 156.32740 days 5812 Tag duration (dive dataset)
## # A tibble: 5 × 6
## id nobs start end dur_days dur_m…¹
## <chr> <int> <dttm> <dttm> <drtn> <dbl>
## 1 22849b 12428775 2014-07-22 13:15:00 2014-12-15 08:29:37 145.84348 days 4
## 2 22850b 11490281 2014-07-17 17:11:00 2014-12-03 11:07:21 138.78913 days 4
## 3 27262 1267945 2013-10-02 13:51:00 2013-10-18 10:39:25 15.86696 days 0
## 4 27262b 11389946 2014-07-22 13:24:00 2014-12-07 12:21:21 137.99816 days 4
## 5 93100 13914413 2014-07-27 07:15:41 2015-01-08 12:49:49 165.27370 days 5
## # … with abbreviated variable name ¹​dur_months2.1 In days
Figure 2.1: Tracking duration (in days) per individual.
2.2 Year of deployment
Figure 2.2: Tracking duration per individual according to year of deployment.
3 Maps of tracks
3.1 Individual maps
Figure 3.1: Individual maps from locations dataset.
3.2 Monthly maps
Figure 3.2: Monthly maps from locations dataset.
There might be a seasonal pattern with the porpoises heading further south from September to the fall. Could also be a simple intra-specific variability (due to different sizes, stages).
4 Bathymetry at locations
4.1 Bathymetry over time
# Mean bathy at locations
summary(loc$bathy)## Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
## 1.0 65.0 109.0 147.3 185.0 1561.0 47# bathy summary per ID
loc %>%
group_by(id) %>%
summarise(mean_bathy = mean(bathy, na.rm=T),
sd_bathy = sd(bathy, na.rm=T),
max_bathy = max(bathy, na.rm=T))## # A tibble: 5 × 4
## id mean_bathy sd_bathy max_bathy
## <chr> <dbl> <dbl> <dbl>
## 1 22849b 208. 194. 1055
## 2 22850b 135. 113. 1561
## 3 27262 166. 94.1 482
## 4 27262b 120. 131. 791
## 5 93100 91.8 78.4 1150# summarize bathy daily per ID
daily = loc %>%
group_by(id, date) %>%
summarise(mean_bathy = mean(bathy, na.rm=T))
Figure 4.1: Mean daily bathymetry (in m) at locations over time for the 5 porpoises.
The individuals globally experienced deep waters (mean: 147 m deep, up to 1561 deep). It is unknown if they would target the bottom, performing demersal dives or rather pelagic dives. One option would be to match the average daily locations to the dive dataset to get an approximation of the mean bathymetry per day, but the different sampling rates between both datasets (1 sec vs. several locs/day) would preclude the investigation of the benthic vs epelagic behaviour in detail.
4.2 Map of bathymetry at locations
Figure 4.2: Map of bathymetry (in m) at locations.
5 Dive depth
5.1 Histogram
Figure 5.1: Individual histograms of dive depth.
The five porpoises performed mainly shallow dives within the first 30 meters, but occasionally conducted deep dives up to 360 m.
5.2 Density
Figure 5.2: Individuals density plots of dive depth.
The histograms show a unimodal distribution except for #27262, for which a bimodal distribution with a peak around 50 m can be seen.
5.3 Profile over time
Figure 5.3: Individuals histograms of dive depth using daily max depth.
These histograms don’t incorporate all the data for each individual, but rather show the daily maximum depth per individual. This plot highlights the deep dives, even though the majority of the dives (short ones) can be somehow masked on these plots. Should look at the proportion of deep dives per individual and per day.
6 Number of dives
# Depth definition: depth threshold=5 m, zoc=0
ndive_summary <- readRDS("./RDATA/1c.diveSummary_5HP_calib_5m_zoc0.RDS")
# number of dives per ID
table(ndive_summary$id)##
## 22849b 22850b 27262 27262b 93100
## 95757 83895 7651 102088 104092# proportion of dives shorter than 1 min (might be underestimated due to the dive threshold of 5 m)
nrow(ndive_summary[ndive_summary$dur<60,]) / nrow(ndive_summary) * 100## [1] 56.62761ndive_summary %>%
group_by(id) %>%
summarise(tot_dives = max(dive),
shorter_1min = n_distinct(dive[dur<60])) %>%
mutate(prop_shorter_1min = (shorter_1min / tot_dives) * 100)## # A tibble: 5 × 4
## id tot_dives shorter_1min prop_shorter_1min
## <chr> <dbl> <int> <dbl>
## 1 22849b 95757 58821 61.4
## 2 22850b 83895 41493 49.5
## 3 27262 7651 2497 32.6
## 4 27262b 102088 59267 58.1
## 5 93100 104092 60742 58.4# proportion of dives below 25 m
nrow(ndive_summary[ndive_summary$maxdep>25,]) / nrow(ndive_summary) * 100 ## [1] 44.10305ndive_summary %>%
group_by(id) %>%
summarise(tot_dives = max(dive),
ndives_below25m = n_distinct(dive[maxdep>=25])) %>%
mutate(prop_below25m = (ndives_below25m / tot_dives) * 100)## # A tibble: 5 × 4
## id tot_dives ndives_below25m prop_below25m
## <chr> <dbl> <int> <dbl>
## 1 22849b 95757 38555 40.3
## 2 22850b 83895 42520 50.7
## 3 27262 7651 4930 64.4
## 4 27262b 102088 40103 39.3
## 5 93100 104092 49768 47.8# proportion of dives below 50 m
nrow(ndive_summary[ndive_summary$maxdep>50,]) / nrow(ndive_summary) * 100 ## [1] 25.95157ndive_summary %>%
group_by(id) %>%
summarise(tot_dives = max(dive),
ndives_below50m = n_distinct(dive[maxdep>=50])) %>%
mutate(prop_below50m = (ndives_below50m / tot_dives) * 100)## # A tibble: 5 × 4
## id tot_dives ndives_below50m prop_below50m
## <chr> <dbl> <int> <dbl>
## 1 22849b 95757 24405 25.5
## 2 22850b 83895 23553 28.1
## 3 27262 7651 2688 35.1
## 4 27262b 102088 22257 21.8
## 5 93100 104092 29947 28.8# proportion of dives below 100 m
nrow(ndive_summary[ndive_summary$maxdep>100,]) / nrow(ndive_summary) * 100 ## [1] 13.34619ndive_summary %>%
group_by(id) %>%
summarise(tot_dives = max(dive),
ndives_below100m = n_distinct(dive[maxdep>=100])) %>%
mutate(prop_below100m = (ndives_below100m / tot_dives) * 100)## # A tibble: 5 × 4
## id tot_dives ndives_below100m prop_below100m
## <chr> <dbl> <int> <dbl>
## 1 22849b 95757 13114 13.7
## 2 22850b 83895 12333 14.7
## 3 27262 7651 1806 23.6
## 4 27262b 102088 9062 8.88
## 5 93100 104092 16590 15.9# dive summary per ID
ndive_summary %>%
group_by(id) %>%
summarise(mean_dep = mean(maxdep),
median_dep = median(maxdep),
mean_dur = mean(dur/60),
median_dur = median(dur/60))## # A tibble: 5 × 5
## id mean_dep median_dep mean_dur median_dur
## <chr> <dbl> <dbl> <dbl> <dbl>
## 1 22849b 41.1 18 1.08 0.6
## 2 22850b 44.5 25.5 1.29 1.02
## 3 27262 62.5 32 1.63 1.5
## 4 27262b 34.8 17.5 1.01 0.667
## 5 93100 48.1 23 1.10 0.683Maximum dives depths were on average between 34.8 and 62.5 m, and 44% of the dives were deeper than 25 m, while only 26% deeper than 50 m. Despite their small proportions, these deeper dives could play a critical role in harbour porpoises feeding ecology. Could these deep dives be related to hunting events instead?
7 Diel pattern
7.1 Max depth vs hours
Figure 7.1: Individual smooth curves of the max dive depth according to the hour of the day.
Based on the maximum daily depth, dives appear to be shallower in the morning (around 8-11 am), as indicated by shallower dives for 4 over 5 individuals. Deeper dives are observed mid-late afternoon (3-6 pm). The same plot should including the entire dataset to avoid bias information based on daily max depth.
7.2 Dive duration
Figure 7.2: Individual histograms of the dive duration according to the phase of the day.
Bimodal distributions during dawn, dusk and night for nearly all porpoises, with dives shorter than 1 min and dives longer than 1-2 min. Dives appear to be globally shorter during the day.
7.3 Max depth
Figure 7.3: Individual histograms of the maximum dive depth according to the phase of the day.
Shallower dives during daytime while bimodality during dusk and dawn.
7.4 Descent rate
Figure 7.4: Individual histograms of the descent rate according to the phase of the day.
7.5 Ascent rate
Figure 7.5: Individuals histograms of the ascent rate according to the phase of the day.
Ascent rate globally higher (faster) than descent rate.
7.6 Bottom duration
Figure 7.6: Individuals histograms of the bottom duration according to the phase of the day.
7.7 All hourly max depth
Figure 7.7: Individuals density plots of the hourly max depth according to the time of the day.
When looking at hourly maximum depth, there is a clear diel pattern in the diving behaviour, with shallower dives during daytime while deeper dives (peak around 150 m) during dawn, dusk and night.
7.8 Individual plots hourly
Figure 7.8: Individuals density plots of the hourly max depth according to the phase of the day.
8 Relationship between max depth and duration
8.1 All ids
Figure 8.1: Relationship between max depth and dive duration for all individuals.
There is an obvious and unsurprising relationship between max depth and dive duration for all individuals.
8.2 Individual plots
Figure 8.2: Individual relationships between max depth dand dive duration.
For all porpoises but #93100, the slope becomes steeper beyond 40 m, meaning that the dive duration will increase less rapidly after a certain depth (approx. 40 m), suggesting the porpoises might reach their ADL. Should look at the relationship between dive duration and post-dive surface interval to investigate that.
8.3 Different periods
Figure 8.3: Relationship between max depth and dive duration for all individuals and different periods.
Slope less steep during daytime.
8.4 Individual plots different periods
Figure 8.4: Individual relationships between max depth dand dive duration.
9 Density plots of max depth vs dive duration
9.1 All dives
Figure 9.1: Individual density plots of the max depth vs duration for all dives recorded.
Except for #27262, most of dives concentrated in shallower waters above 50 m.
9.2 Below 25 m
Figure 9.2: Individual density plots of the max depth vs duration for dives below 25 m.
When looking at dives below 25 m, it seems all dives are still concentrated in the epipelagic layer (around 30 m), with some excursions below 100 m.
9.3 Below 50 m
Figure 9.3: Individual density plots of the max depth vs duration for dives below 50 m.
9.4 Different periods
Figure 9.4: Individual density plots of the max depth vs duration for all dives recorded at different times of the day.
Dives occurring during dawn and dusk seem to be deeper and longer compared to daytime, but the majority of the dives during such periods is still concentrated in shallow layers.
9.5 Dives at dusk
Figure 9.5: Individual density plots of the max depth vs duration for dives at dusk.
9.6 Dives during the day
Figure 9.6: Individual density plots of the max depth vs duration for dives during daytime.
9.7 Dives at dawn
Figure 9.7: Individual density plots of the max depth vs duration for dives during dawn
9.8 Dives at night
Figure 9.8: Individual density plots of the max depth vs duration for dives during nighttime.
9.9 Below 50 m different periods
Figure 9.9: Individual density plots of the max depth vs duration for dives below 50 m at different times of the day.
10 Seasonal pattern
10.1 All ids
Figure 10.1: Boxplot of the maximum daily depth across months.
It seems there is an increasing pattern in terms of max dive depth across months, but it could be related to several factors such as the location of the animal and therefore the bathymetry at such location. Alternatively, could be a real pattern of the porpoises diving deeper during colder months, possibly in relation to the deeper mixed layer depth in winter and its impact of the availability of the prey.
10.2 Individual plots
Figure 10.2: Individual boxplot of the maximum daily depth across months.